Using protein structures predicted by Alphafold2 to understand the fragmentation of proteins from pathogenic bacteria analyzed by MALDI-TOF-TOF-MS/MS

Authors: PARK, JIHYUN - Oak Ridge Institute For Science And Education (ORISE); Fagerquist, Clifton - Keith

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 3/6/2023
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Alphafold2 is an in silico protein folding software/algorithm that is used to predict the structures of proteins. It is now widely used in molecular biology, protein chemistry and drug development. MALDI-TOF-TOF tandem mass spectrometry (MS/MS) and post-source decay (PSD) is a rapid technique that can be used to identify proteins from complex mixtures, e.g. bacterial cell lysates. MS/MS-PSD involves gas phase isolation of specific protein ions prior to fragmentation which allows attribution of fragment ions to a specific protein precursor ion. However, fragmentation efficiency by MS/MS-PSD can vary widely across proteins. We use Alphafold2 to predict the structures of proteins of pathogenic bacteria to better understand their structural properties that can affect protein unfolding and gas phase fragmentation. Alphafold2 was used to predict the structures of bacterial proteins, including: YahO protein; cold-shock proteins CspC and CspE; stress-induced protein CsbD; DNA-binding protein HU; thioredoxin; phosphocarrier protein HPr; immunity proteins of colicin E3 and bacteriocin; a plasmid protein of unknown function, acyl carrier protein; CsgA (a primary component of curli). Experimentally determined post-translational modifications, e.g. removal of N-terminal methionine or signal peptides, were incorporated into the in silico structures. PSD of low charge state intact protein ions is an ergodic dissociation technique that results in polypeptide backbone cleavage on the C-terminal side of aspartic acid (D), glutamic acid (E) and asparagine (N) residues as well as on the N-terminal side of proline (P) residues generating b-type and y-type fragment ions, i.e. aspartic acid effect (AAE). These bacterial proteins showed varying degrees of fragmentation by MS/MS-PSD with some fragmenting efficiently (YahO, CspC, CspE) and others fragmenting very poorly (CsgA and HU). Other proteins showed fragmentation intermediate between these two extremes. The poor fragmentation efficiency of CsgA may be understood due to its compact, highly folded parallel beta-sheet structure stabilized by internal hydrogen bonds with a hydrophobic interior suggesting a strong tertiary structure resistant to protein unfolding and energy deposition in the gas phase. By contrast, the periplasmic protein YahO has a relatively open solvent-accessible structure with two short alpha-helices and a twisting three-strand beta-sheet structure that may facilitate unfolding under the mildly denaturing conditions of the MALDI matrix solution increasing the likelihood of energy deposition by gas phase collisions with matrix molecules during ionization and acceleration from the source. In addition, YahO has seven D-residues whose carboxylic acid side-chains face outward and most are predicted not to have intramolecular hydrogen bonds that might restrict AAE rearrangement and backbone cleavage.